Re: [Gretl-users] missing tests in IV/2sls regression
Doh! I didn't notice I'd left out the intercept.
Re the missing tests, I thought I'd replicated the error on a classroom computer this
morning, only to have them show up as usual when I went to demonstrate the
"bug." I'll check again on my desktop tomorrow.
PS
Sent from my iPad
On Nov 4, 2014, at 6:56 PM, Allin Cottrell <cottrell(a)wfu.edu>
wrote:
> On Mon, 3 Nov 2014, Summers, Peter wrote:
>
> Using mroz.gdt in the PoE 4th ed. data sets, the following works
> as expected in the console or via script:
> tsls l_wage educ exper sq_exper; exper sq_exper mothereduc fathereduc
You really want to run that without an intercept?
> However when I estimate the same model via the gui, the Hausman,
> Sargan and weak instruments tests don't appear (Windows 7, v.
> 1.10.0cvs, 64-bit version).
I've tried it (with and without an intercept) and I'm seeing the
same set of IV tests whether I use a script or the GUI. That's on
Linux. I guess I'll have to try it on Windows.
Allin
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users